Upload
doanthuy
View
217
Download
0
Embed Size (px)
Citation preview
1
To appear in the Journal of Economic Methodology
Method and Appraisal in Economics, 1976-20061
Uskali Mäki Academy of Finland
Ladies and gentlemen, Dr Spiro Latsis,
I am honoured and delighted to address you tonight on this important occasion. I consider
this an important occasion not only for Dr Latsis and this University, but also for the field
of the methodology and philosophy of economics. Honorary doctorates are not abundant
in this field – yet (but I would expect there to be many more in the coming years).
I will begin and end this note with an autobiographical remark.
In the autumn of 1971, I was a fresh undergraduate student, eager to start studies in
economics. After the first few classes of the course on introductory economics, having
been exposed to models and assumptions such as those of perfect competition, my
eagerness took an unexpected twist. I was puzzled enough to conclude that it was
impossible for me to continue without combining my economics studies with my interest
in philosophy. I decided to become a specialist in the methodology and philosophy of
economics. However, at that time, such a research field, as a field, did not yet exist. It
was easy to follow what was published – and because so little was published, it was
relatively easy to catch up, and to read, in the course of a few years, virtually everything
that had been published during the previous two hundred years or so (this would be close
to impossible today given the very different scale of publications in this area and the far
more severe pressures on students). Naturally, I read everything new that was published,
and this was not much in the beginning, thus I was probably among the first to read Spiro
1 Talk given on the occasion of awarding Dr Spiro Latsis an Honorary Doctorate at Witten/Herdecke University, 9 November 2006.
2
Latsis’ 1972 paper in the British Journal for the Philosophy of Science. But the literature
gradually started accumulating. There were Larry Boland’s several articles beginning in
the late 1960s, then his 1982 book on economic method (Boland 1982). There was Alex
Rosenberg’s 1976 book on microeconomic laws (Rosenberg 1976), and Hollis and Nell’s
peculiar 1975 book on rational economic man (Hollis and Nell 1975). There was an
increasing number of journal articles and other contributions of book length, such as
Hutchison (1977), Wong (1978), Stewart (1979), Katouzian (1980), then Blaug (1980),
Hausman (1981), and Caldwell (1982). By this time, the world had changed, and the field
of economic methodology was taking shape as an institutionalized research field.
Of all that was published in the 1970s, it was Spiro Latsis’s edited volume with
Cambridge University Press in 1976 that was to play a major role in setting the research
agenda. Method and Appraisal in Economics became the major source of inspiration for
the next several years. This volume, together with Latsis’s 1972 article that I mentioned,
“Situational determinism in economics” in the British Journal, served to shape the
orientation of the emerging field of economic methodology. Latsis was a student of Imre
Lakatos, the Hungarian born philosopher of science at the London School of Economics,
who took Karl Popper’s ideas closer to the complexities of the history of science. The rise
of Lakatosian economic methodology was to coincide with the rise of economic
methodology as a field. We may therefore say that Spiro Latsis not only helped establish
the Lakatosian project within economic methodology, but that he helped establish the
field of economic methodology itself. Thanks partly to him, there was now a field to
which I had devoted myself before it was there.
To conclude my first autobiographical note, I gratefully admit that even though this
aspiring economic methodologist never became a full-blown Lakatosian nor shared the
Lakatosian research agenda, these works played an important role in my intellectual
development. As I started teaching economic methodology later in the 1970s, I often
included Latsis’s British Journal article in my reader for the students.
3
Three contributions
As I see it, Latsis’s contributions have been at least threefold. He identified “situational
determinism” as the core principle of neoclassical economics and offered a normative
appraisal of it. He introduced Lakatosian principles of describing and appraising
economic theories and programmes. And he played a role as organizer and sponsor of
activities in economic methodology.
Characterisation and Appraisal of Situational Determinism
The first contribution was his work on what he called situational determinism and its
rivals in the context of the theory of the firm (and it was obvious that the significance of
this theme reached beyond the boundaries of firm theory). I think ‘situational
determinism’ is an apt label (more so than Popper’s ‘situational analysis’) for the image
of things conveyed by the traditional neoclassical approach to economic modelling.
According to situational determinism, the situation in which an agent finds itself
determines its behaviour. The inner workings of the agent do not matter, as the external
situation completely constrains its behaviour. The agent’s situation is construed so as to
leave it with no choice: in a single-exit situation, there is just one way to go, so there is
no room for genuine choice, decision-making, and entrepreneurship. Behaviour becomes
a matter of reaction instead of action.
Latsis argued that situational determinism is a degenerating research programme: the
model of monopolistic competition as a successor of perfect competition within the same
programme fails to make even theoretical progress in the sense of generating novel
predictions. Therefore, a case can be made against situational determinism and for an
alternative research programme, economic behaviouralism. Unlike situational
determinism, economic behaviouralism pays serious attention to the psychological and
organisational details of the inner functioning of economic agents. On this programme,
behaviour is not determined by the situation alone, but the inner environment of agents
4
becomes relevant, and behaviour becomes a matter of genuine decision-making in
multiple-exit situations. While the programme of situational determinism is autonomous
with respect to psychology and organisation theory, economic behaviouralism is
dependent on contributions from these fields of inquiry.
Much has happened since the early 1970s in this area. For one thing, monopolistic
competition has made a comeback and, when conjoined with increasing returns, has
turned out to be a fertile source of explanatory force in bringing together fields such as
international trade, urban and regional economics, growth theory and development
economics. A more generous appraisal might therefore be justified today (yet, this may
be complicated by ‘monopolistic competition’ not having been stable in its meaning
through these developments). On the other hand, when presenting and defending
“economic behaviouralism” as a major rival to situational determinism, Latsis’s vision
was just right. We now have “behavioural economics” as a growing and fruitful stream of
economic inquiry that has an identity and label of its own. It combines experimental and
psychological (and recently neuroscientific) work with some traditional themes in
economics, and is indeed interested in the inner workings of economic agents. In the
sphere of the theory of the firm, we now have a plethora of new theories of the firm that
provide accounts of different aspects of firm organisation and behaviour, including
principal-agency theories, transaction cost theories, property rights theories, resource and
capability theories, and so on. Much recent work on firm behaviour is based on crossing
traditional disciplinary boundaries, for example by merging perspectives from economics
with those from law or strategic management. And among other things, entrepreneurial
activity has attracted much more attention than was the case in the early 1970s.
Economic inquiry has moved towards directions that Latsis envisaged and recommended,
but it did so independently of his particular Lakatosian arguments. Yet I would think that
there are still unexploited intellectual opportunities contained in his way of portraying the
alternative images and strategies in terms of (situational) single-exit and (procedural)
multiple-exit modelling. There are some fruit to be collected by intelligent follow-up
5
research framed in these terms, so I would list this as one of the challenges for future
work.
Lakatosian Economic Methodology Introduced and Applied
The developments of Lakatosian economic methodology, on the other hand, did not
proceed independently of Latsis’s contributions. Herein lies his second contribution.
Using the theory of the firm as a case study, he was the first to illustrate the functioning
of the Lakatosian framework by arguing that situational determinism is a scientific
research programme and that it is a degenerating programme. The argument required
framing the issues in Lakatosian terms, those of hard core, protective belt, negative and
positive heuristics, and theoretical and empirical progress. They provide a methodology
of normative appraisal – not prescription – that should meet some descriptive adequacy
requirements. It is this framework that was to dominate large parts of economic
methodology for the next decade and a half or so.
Just to briefly remind the audience of the basics of Lakatos’s (1970) methodology of
scientific research programmes (MSRP), a research programme is a sequence of theories
governed by a set of rules or heuristics. Each theory consists of two kinds of statements,
those constituting the hard core, and those characterising its protective belt. All theories
in the sequence share a common hard core, while each has a more or less different
protective belt. The heuristic of the programme consists of a set of rules that govern the
movements along the sequence. The negative heuristic is the simple proscription against
revising or rejecting the hard core, while the positive heuristic instructs how to modify
the protective belt so as to settle tensions between theory and empirical evidence while at
the same time protecting the hard core. Hence what we get is a sequence of theories
sharing a hard core and governed by positive heuristic for what they do not share, namely
the protective belt. In Latsis’s application, the models of perfect competition and
monopolistic competition belong to the same research programme of situational
determinism. They are members in the sequence of theories that share hard core
assumptions such as profit maximization, perfect information, and independent decisions
6
by firms. They differ little with regard to their protective belt statements: one assumes
homogeneous products, while the other assumes product differentiation. The positive
heuristic of the programme consists of the rules guiding the analysis of equilibrium
conditions and comparative statics.
On the MSRP, theories are held to be empirically accountable, but empirical evidence is
channelled to hit the protective belts only, while leaving the hard core intact. Theory
modification is constrained by empirical evidence, but it is also constrained and guided
by the tenets of a programme so as to guarantee continuity across theory changes. This is
hoped to rule out ad hoc modifications. Yet theory modification is a key idea in the
MSRP. Empirical accountability is not a matter of instant fit between theory and the data;
it is rather a matter of dynamic empirical performance across the sequence of theories.
This is where the concept of progress comes in.
The normative appraisal of a programme is in terms of progress. A progressive
programme is one that exhibits both theoretical progress (it yields a novel prediction each
time there is a move to another theory along the sequence) and empirical progress (those
predictions fail to be contradicted by evidence). A degenerating programme is one that
fails to exhibit theoretical and empirical progress. In Latsis’s application, situational
determinism comes out as a degenerating programme, since monopolistic competition
fails to yield novel predictions in the absence of required input data.
Normative appraisal is essentially comparative. It is comparative in that theory versions
are to be compared across the sequence within a programme. And it is comparative in
that a programme being appraised is to be compared to other programmes for its relative
progressiveness. In principle, degenerating programmes should be refuted and replaced
by progressive programmes. It is well known that Lakatos was unable to solve the
difficult problem of setting rules for determining the conditions under which such
refutation and replacement should happen. For how long (and, I would add, through how
serious failures) is a research programme permitted to degenerate before being
overthrown? The heuristic of a programme would not tell as it is only supposed to govern
7
what happens within the programme. In line with this inconclusiveness in Lakatos’s
framework, Latsis did not rush to radical conclusions in his application. He granted that
the programme of situational determinism may successfully serve as a test bed for the
development mathematical techniques, and that it can be used for answering certain
questions that behaviouralism cannot answer. He was thus unwilling to pass final
judgement in appraising these rival programmes. This was surely understandable also due
to the young age of the behavioural programme.
In the course of the subsequent several years, others published numerous applications of
the Lakatosian framework, including those to international trade, general equilibrium,
new classical macro, and to schools of economic thought such as Keynesian, Marxian,
Austrian, and neo-Ricardian. Few of these studies did what the framework recommended
doing, namely comparing rival research programmes for their relative progressiveness.
They typically examined one programme at a time. It is also noteworthy that even though
Latsis did compare two programmes, he did not compare their rates of progress – only
situational determinism was gauged in terms of progress.
Two Conferences
There is a sense in which the heyday (or even dominance in the field) of Lakatosian
economic methodology lasted for 15 years. And there is a sense in which Spiro Latsis’s
contribution was crucial for both the beginning and the ending of this period. This was in
his role as a sponsor and as an organiser. Method and Appraisal in Economics was
published in 1976 and was based on a conference that took place in the Greek town
Nafplion in 1974 (participants have praised this as the conference of their life time).
Latsis was the organiser of this event that launched the Lakatosian period in economic
methodology. Fifteen years later, in 1989, there was another fabulous conference on the
Italian island of Capri, now sponsored by Latsis who had left the academia. This event
became a collective act of burial of Lakatos in economic methodology, as it were. I was
too young to witness the Nafplion meeting, but I was at Capri, and indeed the atmosphere
was that of a happy funeral. The book based on the conference, Appraising Economic
8
Theories, edited by Mark Blaug and Neil DeMarchi, appeared in 1991, fifteen years after
the publication of Method and Appraisal. But those were important 15 years since they
also were the formative years of contemporary economic methodology as an
institutionalised collective endeavour.
Criticisms of Lakatosian economic methodology
One should not exaggerate: not all work in economic methodology during that period
shared the Lakatosian agenda (such as that of Alex Rosenberg, Dan Hausman, and
myself), and there were criticisms of it from the beginning (indeed, already in Method
and Appraisal by contributors such as John Hicks and Axel Leijonhufvud, and later by
Wade Hands and others). Economic methodology has now more broadly moved beyond
the Lakatosian agenda by addressing new questions about economics and answering them
in novel ways (see DeMarchi 1993; Backhouse 1994a,b, 2008; Davis, Hands, and Mäki
1998; Backhouse, Hausman, Mäki and Salanti 1998; Hands 2001; Kincaid and Ross
2008; Mäki 2008).
Even though the Lakatosian period in economic methodology is over, it is still important
and instructive to be clear about the flaws of the framework. In the remainder of this talk,
I will construe the subsequent developments in economic methodology as a process of
seeking to remove those flaws. This helps see how we got here from there, as the field’s
agenda evolved between 1976 and 2006.
From its early stages, there has been a close alliance between economic methodology and
the history of economic thought. The Lakatosian framework was fit for this alliance given
its focus on the dynamics of progress and its explicit appeal to the verdict of historical
evidence in meta-methodological considerations. Indeed, it is part of the overall
Lakatosian doctrine that methodology should be sensitive or accountable to evidence
pertaining to the history of science. This prepared the way for many contributors to use
historical evidence against the methodological theory. Some other criticisms derived
9
from philosophical and other conceptions of science (see Rosenberg 1986; Hands 1985,
1991; Hausman 1992; Mäki 1990; Backhouse 2008).
Without trying to be exhaustive, we may list the following criticisms against the MSRP:
[1] There is no easily identifiable stable hard core and positive heuristics in economic theorizing. [2] There is no active generation of novel predictions in economics. [3] There is no room in the MSRP for straightforward inductive support by empirical evidence. [4] There is no systematic role in the MSRP for social institutions and their history (they were expelled to the footnotes as Lakatos has it). [5] There is no systematic account in the MSRP of how progress in terms of novel predictions connects with the goal of approaching truths about the real world.
Challenges [1] and [2] complain that economics lacks features that would be required for
the MSRP to apply. Challenges [3], [4], and [5] complain that economics has certain
important features that the MSRP misses. Of these challenges, I find the last three [3]-[5]
as more obviously serious. I will discuss the last two [4]-[5] more extensively after some
brief comments on the first three [1]-[3].
[1] The first complaint, put forth by many, is about the lack in economic theorizing of an
identifiable hard core that would be sufficiently stable across theory developments. Now
without further qualifications, there is possibility of overstressing this matter, and some
of the criticisms may have done just that. The reason for taking this trouble seriously
should not be the fact that the Lakatosian framework simplifies and idealizes. Any
philosophical account – just as any theoretical account in economics and any other
discipline – provides a simplified and idealized image of its subject matter. This would be
reason for a serious charge only if the particular simplifications of an account were
somehow responsible for missing some of the essential features of scientific practice. But
thinking of the notion of a hard core, I find it obvious that in any scientific endeavour,
10
there are more stable and less stable elements, thus it is not obviously dubious to take this
idea close to an idealized extreme in the way Lakatos at times did (and at other times
didn’t). Indeed, Lakatos’s procedure is similar to Marshall’s time period analysis in this
respect: fix the hard core for the short period. On the other hand, in some fields, and in
virtually any field over a long enough time span, instability wins, and the idealization is
defeated. Thus the adequacy of the idealizations becomes an empirical matter that cannot
be decided a priori, but must be considered case by case.
[2] There has been a lot of discussion and debate over the issue of whether economic
theories do, or are able to, generate “novel facts” at all – if they don’t, Lakatos’s
framework will be unfit. This issue is not easy to settle given that there are many notions
of novelty and that Lakatos himself held several of them (Hands 1991). It does seem
obvious that a move ahead along a sequence of economic models does not
characteristically help predict some entirely unheard-of facts, and also that the facts to be
explained or predicted often play a role in motivating the construction of those new
models – thus there is no novelty in these senses. But other kinds of novelty may be
attained, such as a skilfully invented novel way of deriving familiar phenomena from the
premises of standard theory. What would be novel is not some predicted phenomenon P,
but rather the derivability of P from theory T. This is generally viewed as an important
achievement in economics, and it surely qualifies as scientific progress of some sort. But
it is less evident how weighty a role it should play in theory appraisal. More work is
needed on this issue.
[3] The Popperian-Lakatosian denial of the rationality of positive evidence providing
inductive support to scientific theories is a peculiarity that many commentators have
found unacceptable. Everyone grants that inductive inference is not deductively valid, but
this does not have to be taken to imply that it is never rational. Induction, in many guises
and in many locations of the endeavour of science, seems to be among the essential
features of successful scientific practice, thus dismissing it seems descriptively and
normatively inadequate. More work is needed to identify and assess actually practiced
inductive inference in economic inquiry.
11
In the remainder of this talk, I’ll pay my main attention to the last two complaints:
[4] The secondary or tertiary and unsystematic role (footnote role at most) given to social
history and institutional structure in appraising a theory or programme is unsatisfactory in
contemporary lights. The social aspects of science do not belong to footnotes only.
[5] Realists – such as the followers of Popper and Lakatos – should wish to show a
systematic connection between dynamic empirical adequacy and increasing truthlikeness.
They should show how the connection flows from truth to empirical adequacy rather than
the other way around. This is missing in Lakatos’s account.
Lakatosian MSRP as a stage in the progress of economic methodology
The remarks to follow discuss issues [4] and [5] and suggest construing Lakatos’s
account as a progressive stage in the development of the philosophy of science and of
economics in particular. These observations deal with issues of appraisal under the
headings of Unit of Appraisal – Method of Appraisal – Target of Appraisal – Approach
to Appraisal.
Unit of Appraisal
The line of progress in our methodological understanding of the appropriate unit of
appraisal can be simplifies as follows:
Single hypothesis Research programme Institutional structure
Here the idea is that Lakatos expanded the unit of appraisal from a single hypothesis to a
larger research programme, and that this was an improvement in our understanding of
science, economics included. However, he did not go far enough by including a
systematic account of how to normatively appraise the social (academic and otherwise)
12
institutions that shape the developments of theories and programmes. Before him,
Thomas Kuhn had combined ideas akin to programme and institution in his notion of a
paradigm – but Kuhn lacked an idea of rational appraisal when dealing with paradigms as
such cognitive-institutional conglomerates. Here Lakatos and Kuhn had better be
reconciled: everything in and about science, from the cognitive to the social, should be
made subject to rational and critical appraisal.
There is another reconciliation that seems obvious when considering the unit of appraisal
issue. Instead of choosing one kind of unit in exclusion of others as a general principle,
the choice should be contextual: depending on the situation and the task at hand, the
appropriate unit may be this and it may be that – single hypothesis, model, theory, larger
programme, or institutional structure. All sorts of units have to be assessed, and their
assessments do not substitute for one another. In appraising a single hypothesis, it is not
sufficient to check whether the institutional structure is appropriate or whether the
research programme is progressive – just as in appraising the credentials of a programme
or an institutional structure, testing a single hypothesis is not enough. There is no single
and fixed all-purpose unit of appraisal in science.
Method of Appraisal
In regard to the method of appraisal, the line of progress proceeds as follows:
Instant test of an isolated hypothesis Appraisal of the comparative dynamics of research programmes Appraisal of epistemically relevant characteristics of the institutional structures enabling and constraining inquiry
Again, I take Lakatos to have helped make progress in our understanding of the
appropriate method of appraisal. Instead of checking isolated hypotheses against
empirical evidence at an instant of time, science should estimate the comparative
performance of research programmes over time in terms of the theoretical and empirical
progress they are hoped to exhibit. And again we must go beyond Lakatos, since there is
further meta-methodological progress to be made by insisting that the institutions of
13
inquiry be tested and evaluated as well. Institutional appraisal is a matter of epistemic
evaluation of the relevant rules and norms, conventions and customs, incentive structures
and reward systems, the rhetoric and ethics of communication -- the industrial
organisation, constitutional structure and geopolitics of economic inquiry, including its
institutions for education, research, publishing, public relations, and connections with
policy advice. These are epistemic institutions that require epistemic evaluation for their
epistemic consequences. Here is a way of putting the reason why institutions must be
appraised: research programmes are created and appraised by people acting and
interacting under various institutional constraints; the existence and contents of these
programmes and the reliability of these appraisals may thus crucially depend on those
institutions; therefore, those institutions must themselves be appraised for their qualities
in enhancing the innovative production and reliable appraisal of theories and programmes.
Among other things, this might help towards remedying an awkward implication of
Lakatos’s MSRP: namely, any research programme receiving too little attention –
because it is examined by nobody, or too few, or too incompetent, or too low-status
scientists – will automatically be judged as degenerating. Attention, competence, prestige,
and numbers are institutionally dependent matters, and they, too, must be normatively
appraised. The appraisal of these matters cannot be derived from the results of Lakatosian
appraisal, because the reliability of the latter is dependent on just such institutional
conditions being right.
All this is to advocate the normative branch of social epistemology as we know it today.
In contrast to the sociology of scientific knowledge that has a descriptive and explanatory
orientation, the agenda of social epistemology also includes the task of normative
appraisal (cf. Mäki 1992, 2005; Hands 1994). Social epistemology as a normative inquiry
into the social conditions of knowledge acquisition has become a major approach to these
issues in the last two decades or so – that is, after the 1970s contributions by Lakatos and
Latsis. It is noteworthy that this advocacy does not mean a return to Kuhn’s thought that
was so strongly resisted by Popper and Lakatos: in Kuhn’s account, the institutional
structure involved in a paradigm is not normatively appraised, it is rather taken as an
14
unquestioned given. In a sense, then, the social epistemology endorsed here is non-
Kuhnian and retains a Lakatosian spirit of normative appraisal. On the other hand, it is
non-Lakatosian in lifting the social aspects of inquiry from the footnotes to the core text
itself.
Target of Appraisal
For what qualities are the units of appraisal appraised? Any given hypothesis, research
programme, and institutional structure has a vast number of properties, but only some of
them are relevant to epistemic appraisal. As I see it (and here my personal philosophical
convictions start playing a growing role), the line of progress goes as follows:
Static empirical adequacy Dynamic empirical adequacy Truthfulness / truth conduciveness
Theories, research programmes and so on are being appraised, assessed, evaluated – but
precisely for what, for which of their properties? It is not enough to say that they are to
be appraised for their “epistemic value” as this is not a very specific property as such.
One has to add some idea about what epistemic value consists in. It is here that one’s
philosophical outlook starts making a difference. In the empiricist tradition, epistemic
value is a matter of empirical fit, consistency with empirical evidence, or generally
empirical adequacy. Static empirical adequacy can be measured in terms of an instant
empirical test, mentioned above when discussing the method of appraisal. Progress in the
Lakatosian sense is a dynamic form of empirical adequacy, measured in terms of the
empirical performance of a programme over time. It is a matter of a research programme
generating corroborated “excess content” across the relevant sequence of theories –
where content is understood in terms of the predictive implications of a theory.
But we may want to go further than this, and ask: what does dynamic empirical adequacy
measure, what further quality does it manifest? Lakatos himself added presumptions to
the effect that progress in dynamic empirical adequacy is accompanied by progress in
getting closer to the truth, but this connection remained unanalyzed. It also seems
15
obvious that for him, just as for Popper in the latter’s failed doctrine of verisimilitude, the
acquisition of truth was a matter of getting closer to a detailed, accurate, and complete
description of concrete phenomena. But this idea can be questioned. Besides complete
and detailed truth, there may be incomplete truths, that is, truths about some limited yet
significant aspects of the world – such as distinct causal mechanisms. The Lakatosian
framework seems unfit for dealing with such truths (Mäki 1991). In order to have an
account of the possibility of making progress in economics by way of increasing
truthfulness or truth conduciveness, one needs a refined account of the structure of
economic theorizing and modelling – more refined than one finds in Lakatos. Such an
account must be careful with the various functions served by the idealizing assumptions
in models as well as with the intended truth bearers when using those models to acquire
information about the world. In such an account, empirical adequacy plays a criterial role
rather than being the ultimate target of appraisal. These issues are on the agenda of
contemporary economic methodology. (Mäki 2004).
In this connection, it may be useful to compare the above ideas to the project on the
rhetoric of economics – another development that has challenged the Lakatosian
methodology of economics (e.g. Klamer, McCloskey, and Solow 1988) Let us first note
Lakatos’s (and Popper’s) anti-psychologism: the epistemic value of a theory or research
programme is an entirely objective matter and does not depend on mental matters such as
whether any particular individual or group – or anybody at all - believes it, is committed
to it, or even understands it. Now compare this to the much debated conception of
economics as rhetoric, the idea that the communication amongst economists and others
amounts to persuasion of audiences. On this view, persuasiveness is the key property of
theories and research programmes that is to be pursued by economists. Economists seek
to persuade their audiences by various means in an ongoing conversation, and scientific
success is a matter of succeeding in the attempt to persuade. Now, being persuaded and
persuasiveness are notions that are best understood in terms of belief: persuasion is
modification or manipulation of belief. Thus these notions are among the non-objective
mental matters that are abhorred by Lakatosian anti-psychologism.
16
In my view, persuasiveness alone cannot constitute a sensible ultimate target, and cannot
itself constitute epistemic value. But there is the further possibility that, on some stringent
conditions. it plays a criterial role in assessing the epistemic value of a theory or research
programme (Mäki 1995). We know that falsehoods are often very persuasive, marketed
as they are by skilful persuaders such as politicians, commercial advertisers, sometimes
scientists. Therefore, indeed, in order for persuasiveness to serve a criterial purpose in
epistemic appraisal, we must require that certain stringent conditions be met. Again, these
conditions are largely a matter of the institutional structure of the relevant epistemic
communities, including the ethics of communication. Hence, even though truthtfulness
were regarded as the final target of epistemic appraisal, indicators and criteria – such as
empirical adequacy and persuasiveness - are needed for measuring how well the target
has been captured or approached. And the reliability of those criteria depends on the
adequacy of the institutional conditions for the pursuit of epistemic goals. Therefore,
those institutional conditions themselves become included in the relevant set of criteria.
Approach to Appraisal
The issue here is about how precisely we should view the relationship between truth and
other qualities of economic theories in one’s conception of progress.
Progress as increasing empirical adequacy, with a wish of thereby getting closer to the truth Progress as tracking truths about the real world, with (progress in) other qualities of theories as indicators and facilitators of this achievement
This is related to the previous issue of targets: what is the direction of dependency
between empirical adequacy (and other qualities, such as persuasiveness or plausibility)
and truth, and how are we to approach the examination of their relationship? On
Lakatos’s account, we begin with evidential demands on research programmes, then we
postulate – but only postulate - that meeting those demands will have a chance to take us
closer to the truth, too. The alternative view is more strongly realist in outlook,
suggesting that we begin with demands on truth. What it is for a model to be true, given
certain explanatory interests? What is it for inquiry to make progress in the sense of
17
tracking truths about the world? And we may say, as one possibility among others, that
for a model to be true, it must help reveal a causal mechanism inside what used to be a
black box. After this, we ask: what other kinds of progress may enhance meeting these
demands, or help us identify moves towards meeting them? In order to answer this latter
kind of question, we need ideas about further kinds of progress that facilitate truth
tracking and those that serve as indicators of the extent to which truth tracking is
successful.
The facilitators of truth tracking include things such as questions, techniques, and
institutions. Progress in questions may be a matter of raising new kinds of questions or
raising sharper versions of old questions, and this may help give answers that track new
kinds of truths or sharper truths. Progress in techniques may do many things, such as help
render tractable important problems and questions that otherwise may have been difficult
or impossible to answer; it may also be a matter of making available new and more
reliable bodies of evidence. Progress in institutions enables and encourages innovation
and critical conversation, and these help track new truths and eliminate harmful
falsehoods and other mistakes.
On the other hand, among kinds of progress in the indicators or criteria of truth tracking
we typically include empirical advancement. This may be taken to be a matter of many
kinds of thing, such as deriving novel predictions, increasingly accurate predictions, or
more broadly applicable predictions, from one’s theory. In considering whether to accept
increases in the degree of consensus or persuasion in the relevant research community
among the criteria or indicators, we must, to repeat, make their reliability dependent on
the extent to which the institutions of inquiry are right.
The above suggestions imply that our methodological theories of the appropriate criteria
of appraisal are themselves subject to change and progress. Lakatos permits progress in
the theories about appraisal criteria, but he believed himself to have articulated the
correct theory of those criteria (captured by his notions of theoretical and empirical
progress, to be applied to developments within a programme). Lakatos also permits
18
progress in the application of his favourite criteria in scientific practice (for example, he
might have conceived of these applications becoming increasingly consistent and
reliable). All this is fine as far as it goes, but I have argued above that this image of things
is too narrow, therefore we need a broader array of kinds of progress that are linked to
one another in a variety of ways (as facilitators and indicators). By broadening the array
of effective kinds of scientific progress we ourselves make methodological progress.
We might also consider taking a step further. On Lakatos’s view, the correct criteria of
appraisal are supposed to be universal and fixed across our theories and applications of
them. In contrast, we may need an idea of progress in the criteria of appraisal themselves.
As science changes and is done in changing institutional conditions, we should be
prepared for the possibility that the appropriate criteria of appraisal also change. This
creates another possibility for making methodological progress.
I should separately warn against being too narrow in the other direction. This inadvisable
trend could be characterised as becoming overimpressed by the institutional shaping of
scientific practice and giving up the very idea of normative appraisal. While Lakatos
downplayed the importance of institutional realities, the reverse trend has now become
popular in the metatheoretical study of economics, shifting attention away from rules of
appraisal of the sort we find in Popper and Lakatos. But we should not forget that much
of scientific practice deals with how theories are linked to empirical evidence, therefore
economic methodologists also need to reintroduce the Lakatosian focus on this level. My
proposal is to pursue methodological progress by combining descriptive, explanatory, and
normative perspectives to economics while acknowledging the multiplicity and
complexity and interconnectedness of the various kinds of conditions and criteria, from
cognitive to institutional, that shape – and should shape – economic inquiry. Put in terms
of appraisal, a simplified version of my proposal goes: appraise theories and
explanations for their truth by appraising their empirical performance and by appraising
the institutional conditions of theoretical and empirical inquiry. As I see it, Lakatos’s
MSRP seriously deals with issues of empirical performance only, whereas our horizons
19
should be expanded so as to encompass issues of truth and institutions as well (while
Lakatos considered the former by postulation and relegated the latter to footnotes).
One way of formulating some of these ideas is to suggest that the “naturalisation” of
economic methodology is to be taken a few steps further. Lakatos’s methodology is
naturalistic in the sense of being made accountable to empirical evidence pertaining to
the actual history of science. The next step, already being taken by many, is to make the
contents of the standards used in methodological description, appraisal, and prescription
dependent on the contents of the best scientific theories about the structure and
functioning of the cognitive agents and relevant epistemic institutions (cf. Hausman 2001;
Hands 2001; Mäki 1992, 2005). The contents of methodology, in other words, should
interact with the contents of fields such as cognitive science, sociology of science, and
the economics of organisations and institutions.
Let me support this idea with an analogy to an argument presented some three decades
ago. The argument was presented by Spiro Latsis in 1972 and again in 1976, and I have
already summarised it above. Latsis argued that economic theory based on the
programme of situational determinism was to be improved by making it less autonomous
and closed with respect to research fields dealing with the structure and functioning of
economic agents – such as psychology and organisation theory. In analogy to Latsis’
argument, economic methodology is to be improved by making it less autonomous, by
welcoming influences from similar substantive research fields so as to enrich our image
of real scientific agents in action. What we need is a behavioural and institutional
economic methodology to highlight the complex mental and social processes that shape
the development of economic inquiry.
[After having delivered this address, Chris Mantzavinos gave me a copy of a brief paper
by Spiro Latsis, published in 1996 in a Festschrift for Axel Leijonhufvud. Entitled
“Nafplion revisited”, the paper surprised and delighted me by suggesting the same
analogy between situational determinism and “rationalist” accounts of science, and by
emphasizing the need to go beyond them by examining the causally influential ways in
20
which internal mental factors and external social factors shape inquiry especially in the
social sciences. Even though busy with other business – or perhaps thanks to it? – Dr
Latsis keeps making sharp and up-to-date observations about Life Among the Econ, to
borrow Leijonhufvud’s expression.]
Conclusion
The Lakatosian stage in the progress of the philosophy and methodology of economics
has served a useful role in helping launch a systematic and collective field of inquiry. In
the course of the thirty years between 1976 and 2006, the field has advanced beyond that
stage, but much of the spirit of the original ambitions had better be retained. As much as
ever, if not more than ever, we need economic methodologies for normatively appraising
assumptions, models, theories, explanations, research programmes, and, importantly, for
appraising academic and other institutions for their capacities in helping us acquire truths
about the real world. Given the increasing complexity, diversity, and fragmentation of
economics itself, practicing economists should be increasingly receptive to
methodological reflection.
I would like to congratulate Universität Witten/Herdecke for making the wise decision of
awarding Spiro Latsis an honorary doctorate. I began this talk with an autobiographical
note, and promised to finish it with another. About ten years ago, soon after my arrival at
Erasmus University of Rotterdam, I made a formal proposal to nominate Spiro Latsis for
an honorary doctorate at Erasmus University. The proposal was not accepted by those
who had the power to decide. So Dr Latsis had to wait for another decade - and another
place - to receive the well deserved title. Congratulations.
21
References Backhouse, Roger, ed. (1994a) New Directions in Economic Methodology. London: Routledge. Backhouse, Roger (1994b) “The Lakatosian legacy in economic methodology”, in New Directions in Economic Methodology, ed. R. Backhouse. London: Routledge. Pp. 173-191. Backhouse, Roger (2008) “The rise and fall of Popper and Lakatos in economics”, in Handbook of the Philosophy of Economics, ed. U. Mäki. Elsevier. Backhouse, Roger, Daniel Hausman, Uskali Mäki, and Andrea Salanti, ed. (1998) Economics and Methodology. Crossing Boundaries. London: Macmillan. Blaug, Mark (1980) The Methodology of Economics. Cambridge: Cambridge University Press. Blaug, Mark and Neil DeMarchi, ed. (1991) Appraising Economic Theories. Studies in the Methodology of Research Programs. Aldershot: Edward Elgar. Boland, Lawrence (1982) Foundations of Economic Method. London: Allen & Unwin. Caldwell, Bruce (1982) Beyond Positivism. Economic Methodology in the Twentieth Century. London: Allen & Unwin. Davis, John, D. Wade Hands, Uskali Mäki, ed. (1998) The Handbook of Economic Methodology. Cheltenham: Edward Elgar. DeMarchi, Neil, ed. (1993) Post-Popperian Methodology of Economics. Dordrecht: Kluwer Publishers. Hands, D. Wade (1985) “Second thoughts on Lakatos”, History of Political Economy 17, 1-16. Hands, D. Wade (1991) The problem of excess content: Economics, novelty, and a long Popperian tale”, in Appraising Economic Theories, ed. M. Blaug and N. DeMarchi. Aldershot: Edward Elgar. Pp. 58-75. Hands, D. Wade (1994) “The sociology of scientific knowledge”, in New Directions in Economic Methodology, ed. R. Backhouse. London: Routledge. Pp. 75-106. Hands, D. Wade (2001) Reflection without Rules. Economic Methodology and Contemporary Science Theory. Cambridge: Cambridge University Press.
22
Hausman, Daniel M. (1981) Capital, Profits, and Prices. An Essay in the Philosophy of Economics. New York: Columbia University Press. Hausman, Daniel M. (1992) The Inexact and Separate Science of Economics. Cambridge: Cambridge University Press. Hausman, Daniel M. (2001) “A new era for economic methodology”, Journal of Economic Methodology, 8, 65-68. Hollis, Martin and Edward Nell (1975) Rational Economic Man. Cambridge: Cambridge University Press. Hutchison, Terence (1977) Knowledge and Ignorance in Economics. Oxford: Blackwell. Katouzian, Homa (1980) Ideology and Method in Economics. London: Macmillan. Kincaid, Harold and Don Ross, ed. (2008) Oxford Handbook of the Philosophy of Economics. Oxford: Oxford University Press. Klamer, Arjo, Donald McCloskey, and Robert Solow (1988) The Consequences of Economic Rhetoric. Cambridge: Cambridge University Press. Lakatos, Imre (1970) “Falsification and the methodology of scientific research programmes”, in Criticism and the Growth of Knowledge, ed. I. Lakatos and A. Musgrave. Cambridge: Cambridge University Press. Pp. 91-196. Latsis, Spiro J. (1972) “Situational determinism in economics”, British Journal for the Philosophy of Science. 23, 207-245. Latsis, Spiro J., ed. (1976) Method and Appraisal in Economics. Cambridge: Cambridge University Press. Latsis, Spiro J. (1996) “Nafplion revisited”, in Inflation, Institutions and Information. Essays in Honour of Axel Lijonhufvud, ed. D. Vaz and K. Velupillai. London: Palgrave Macmillan. Pp. 24-32. Mäki, Uskali (1990) "Methodology of economics: Complaints and guidelines", Finnish Economic Papers, 3, 77-84. Mäki, Uskali (1991) "Comment on Hands", in Appraising Economic Theories, ed. N. DeMarchi and M. Blaug. Aldershot: Edward Elgar, 1991. Pp. 85-90 Mäki, Uskali (1992) "Social conditioning of economics", in Post-Popperian Methodology of Economics, ed. N. DeMarchi. Dordrecht: Kluwer Publishers. Pp. 65-104.
23
Mäki, Uskali (1995) "Diagnosing McCloskey", Journal of Economic Literature, 33, 1300-1318. Mäki, Uskali (2004) “Some truths about truth for economists, their critics and clients”, in Economic Policy-Making under Uncertainty: The Role of Truth and Accountability in Policy Advice, ed. P. Mooslechner, H. Schuberth, and M. Schurtz. Cheltenham: Edward Elgar. Pp. 9-39. Mäki, Uskali (2005) “Economic epistemology: Hopes and horrors”, Episteme. A Journal of Social Epistemology, 1, 211-220. Mäki, Uskali, ed. (2008) Handbook of the Philosophy of Economics. Elsevier. Rosenberg, Alex (1976) Microeconomic Laws. Pittsburgh: Pittsburgh University Press. Rosenberg, Alex (1986) “Lakatosian consolations for economists”, Economics and Philosophy 2, 127-139. Stewart, Ian (1979) Reasoning and Method in Economics. An Introduction to Economic Methodology. London: McGraw-Hill. Wong, Stanley (1978) The Foundations of Paul Samuelson’s Revealed Preference Theory. A Study by the Method of Rational Reconstruction. London: Routldge & Kegan Paul.